In late November, joint IQE-Cardiff University venture, the Compound Semiconductor Centre, bagged just over £350,000 in UK government funds to develop ultra-sensitive magneto-sensors with integrated electronics based on GaAs and GaN materials.
The InnovateUK award to the new CS-MAGIC project comes at a time when industry analysts predict astonishing growth in the magnetic field sensor market. MarketsandMarkets forecasts that today's $3 billion global market will mushroom to more than $5 billion by 2023, with a compound annual growth rate of 8.77%. Similarly, Technavio forecasts a 10% CAGR from now until 2020.
Right now, demand is largely coming from the automotive industry, requiring highly sensitive, robust Hall Effect sensors to map ever-smaller magnetic fields across larger temperature ranges in engine control management, anti-lock braking systems and more. However, the sensors are increasingly demanded in aerospace and industrial sectors to measure rotation, speed and linear position.
As Mohamed Missous, Professor of Semiconductor Devices and Materials at the University of Manchester and one of the award recipients, puts it: "We're now developing sensors for harsher environments than ever before."
"When today's commercially available silicon Hall Effect integrated circuits reach 150ºC, they just stop working, so devices made of III-V semiconductors are attracting a great deal of interest," he adds.
What's more, conventional silicon-based devices can also suffer from low magnetic field sensitivity, limited operating frequency and temperature ranges, and high power consumption.
So, for his part, Missous has been developing Quantum-Well Hall ICs based on GaAs-InGaAs-AlGaAs systems, having also founded UK-based Advanced Hall Sensors (AHS) to take his 2DEG Hall sensors to market.
Devices are fabricated using molecular beam epitaxy to deposit the thin films onto GaAs wafers, which are then processed into Hall sensors via subsequent lithography, etching, annealing and metal evaporation steps before being sliced and packaged.
Crucially, quantum wells are generated by inserting a thin InGaAs film between two thicker AlGaAs layers. Electrons - in the form of a two dimensional gas or 2DEG - are confined within this thinner layer, rather than being left to roam free in conventional Hall Effect sensors, boosting electronic properties.
Indeed, silicon-based linear Hall ICs can detect magnetic fields down to around 600 nanoTesla in a 10 Hz bandwidth. But AHS has already produced quantum well Hall 2DEG GaAs-based devices that detect magnetic fields as low as 177 nT in a 10 Hz bandwidth. And while the former devices have a maximum cut-off frequency of 10 kHz, the wide bandgap versions stretch this figure to 200 kHz.
To date, AHS has already shipped more than 15 million discrete sensors, but thanks to the latest funds, Missous and his company now intend to commercialise single-chip sensors with a wider dynamic range that can operate in harsher environments.
"Our all-integrated chip will be a single chip that has the quantum well Hall effect sensor in it with all the drive electronics," says the researcher. "This will be totally new; a single chip GaAs Hall Effect sensor just doesn't exist right now."
The drive electronics will also be based on GaAs, which means the entire chip will be more radiation resistant and will operate across a wider temperature range than existing devices. Indeed, on temperature range, Missous is aiming for -200ºC to 200ºC with the single chip, while today's silicon device cover -50ºC to 150ºC.
Right now, AHS is working with industry partners, TWI and Renishaw, to develop a generic structure based on GaAs-InGaAs-AlGaAs layers. This should be confirmed come April next year, and by the following November, the sampling of packaged devices should be underway.
"We want to produce generic ICs that operate over the higher temperature range and once we have achieved this, TWI and Renishaw will test the devices and feedback on performance," says Missous. "We are relying on the companies to do this so we can tailor the chip to whatever they require."
"We do not want a very complicated circuit but we need to ensure that all the electronics are in sync with the sensor," he adds. "The circuit will work over enormous temperature ranges without losing any sensitivity or functionality across these."
Raising the bar
Meanwhile, in a second strand of the CS-MAGIC project, Dr Petar Igic from the Electronic Systems Design Centre at the University of Swansea is developing what could be described as the next wave of magnetic sensing based on GaN.
Igic has fabricated GaN HEMTs for power applications, but now, with the latest InnovateUK funds, intends to develop discrete sensors based on his magnetic HEMT concept. "Our aim has been to develop a sensor that is fully compatible with existing HEMT technology," he says. "We don't need different starting materials or extra processing steps but the layout is different to the HEMT so that the device is sensitive to magnetic fields."
Crucially, GaAs-based sensors can operate up to 220ºC, but sensors based on GaN HEMT structures could maintain performance up to around 400ºC, subject to, as Igic puts it, 'the development of adequate packaging and soldering'.
Right now Igic is working on proving the concept and optimising device layout, which he hopes will be achieved within the next twelve months. After this, he will be looking to tailor the device for specific applications, and will also be working with packaging companies
Clearly,CS-MAGIC's GaAs-based Hall sensor is closer to commercialisation than the second, GaN-based device. But the potential performance gains of GaN in harsh environment applications are undisputed.
According to Igic, initial results indicate that the GaN-based sensor detects magnetic fields as low as 100 nT. What's more, device sensitivity will not change with magnetic fields and high temperatures.
"We cannot beat silicon on price but this GaN device will reach aerospace and automotive applications as it will work at very high temperatures and is less sensitive to radiation," highlights Igic. "We will have to implement the device in standard manufacturing processes, but its unique selling point is that when GaAs stops working, this still is."